Can someone with knowledge on US Chess FIDE policy help clear this up for me.
Is it acceptable or unacceptable for a pairing program to call itself ‘FIDE Compatible’ and provide Tournament Directors with a way to spoof SwissSys files so they can submit tournaments to the USCF as dual rated if that program uses an approved engine (e.g. bbpPairings, JaVaFo, etc.) under the hood despite the program itself never going through the FIDE channels to become approved software?
I always thought the answer to this was no. I reached out to someone at US Chess months ago and they told me it that was not acceptable as well. But since then, it sounds like the office may be telling other people this is acceptable. I won’t mention the specific pairing program that is actively doing this today or the club that is running dual rated events with this uncertified software, but what would happen to those events (or should happen) if this came to light?
My understanding is that if you want to call a pairing program “FIDE compatible”, you must get FIDE to certify it, and, as with nearly anything involving FIDE, there’s a fee associated with that.
You can probably note that it can prepare a file using the FIDE ratings format (aka the Krause format) without FIDE permission. Whether that file is sufficient for the US Chess office to submit a rating report to FIDE is a separate question.
It’s not, as the US Chess requirements currently state that “US Chess requires that all FIDE rated tournaments be submitted to US Chess for processing using a FIDE Endorsed Pairing Program.” but in practice mean “send us a SwissSys tournament file”. I haven’t heard of anyone being able to send US Chess Vega or Tornelo files, for instance.
Specifically, at the end, where it says “Usage of Equipment at Tournaments (See General Regulations for Competitions).”
For non-norm granting tournaments (FIDE Level 3), it says:
8.3.1 Endorsed and Accepted Products are recommended. Self-Declared Compliant Products are tolerated.
8.3.2 The Chief Arbiter can authorise the use of Equipment lacking the prescribed acceptance level. See General Regulations for Competitions under General Rules and Technical Recommendations for Tournaments.
So, the Chief Arbiter can choose to authorize the use of software that is self-declared compliant. There are further details on what self-declared compliance means in their documentation.
I’m not sure if the office staff even have copies of those programs or some of the others that have popped up in the last decade. Tom Doan told me a year or so ago that he was hoping WinTD would have FIDE certification soon, I don’t know if that has happened.
What the office (or specifically Brian Yang) does with the files is probably the key. I haven’t looked at the Krause format recently, but I had several discussions with him during and after the Turin Olympiad and as I recall one of the issues I had with the Krause format was that it required the date of each game (probably so FIDE can make sure rules regarding how many games/hours players spend at the board each day are not violated), something that was not available on a crosstable and which some pairing programs might not even capture.
The files get submitted to the US Chess office though and as @ulmont pointed out, they have clear requirements. It isn’t enough to self declare your compliant and then spoof SwissSys files to obfuscate the source.
There is actually a reason that SwissSys files are the only ones accepted. Of the US centric pairing programs, it is the only one that is approved.
I don’t mind being direct that you’re referring to my software.
I’m certainly not trying to “spoof” anything. The files say the pairing program name in them in the part of the file format that says what program they came from.
My understanding was that this is a file format that enables a process. They can be opened in SwissSys, and SwissSys can do all of the right things to get the right data for submission to FIDE and processing at USCF. I’ve sent some test files to Brian to make sure they could be opened and processed appropriately.
My point above covers what is considered allowed to FIDE. The bar is pretty low if it’s not an event granting norms.
One other note, we’ve already successfully gone through this process end-to-end with a USCF+FIDE rated event. I’m happy to adjust in any way necessary, though! I just want to make it easy on everyone involved.
Until March of this year, FIDE had an endorsement program where FIDE “recommends” (old C.04 Appx A) that certain programs should be used, and tournament directors were given the latitude to choose other non-endorsed programs provided the TD had confidence that the pairings generated would conform with endorsed programs. Compliance with the pairing rules is the ultimate requirement, after all, as we could theoretically pair tournaments by hand.
Given that the pairings need to be rules conformant, there are multiple open-source engines that can perform Dutch Swiss systems, including JaVaFo and bbpPairings, and those are the core of multiple FIDE-endorsed programs (and a reason FIDE identifies the engines when listing its recommended software). As a TD, when a non-endorsed program uses those same pairing engines it gives me confidence that they produce the right pairings too. In one case of TournaChess, which uses bbpPairings as its pairings engine, we simulated events and cross-compared with SwissSys, and indeed they produced identical results. We’ve double-checked real events against SwissSys too, with no issues. We’ve run a TournaChess event, checked it with SwissSys, submitted it to USCF and had a tournament successfully FIDE rated this way through USCF with no issues.
The recent changes to the FIDE rules seem to tighten the requirements on FIDE norm tournaments and major international events, that the software used have varying review/approvals by FIDE. But FIDE preserved the old standards for Level 3 (non-norm) events that endorsed programs are recommended through the language of B.01 “8.3.2 The Chief Arbiter can authorise the use of Equipment lacking the prescribed acceptance level. See General Regulations for Competitions under General Rules and Technical Recommendations for Tournaments” meaning that any pairing program can be used with the Chief Arbiter’s approval – even if not endorsed or even self-declared compliant.
In the end, I think the more pairing programs the better. SwissSys and WinTD are relics compared to what can be achieved nowadays. I’d like to see ChessNut and other programs use JaVaFo or bbpPairings to achieve the same functionality! The more innovation/variety/competition in chess software the better!
I’ve registered as a “manufacturer” with FIDE. The SDPC bit is in progress, pending a response to a problem I’m having with their forms.
@brandon 's message helps clarify the rules further though. I started based on my understanding of the pre-March rules. Based on my read of this March update, I’ve been trying to move up a rung to SDPC (and eventually higher, whenever that feels like an appropriate step for what we’re doing), though the formal step does not appear to be strictly required.
I agree that it would be nice if the requirements were looser. Chess Nut would include bbpPairings. In fact, the engine is already there but it is not enabled for US Chess clubs because of the clear US Chess guidelines:
US Chess requires that all FIDE rated tournaments be submitted to US Chess for processing using a FIDE Endorsed Pairing Program.
If that is contradicted by US Chess elsewhere or is since stale information, I’d love to know. That way we can confidently turn it on for folks in the US domain.
and maybe we can collaborate on a way to support USCF on an easy and automated way to validate that files that programs like ours generate include all necessary information. I want this to be easy on the USCF side, too!
I could also share more details on the file format if it’s going to turn into a de facto standard for how we submit USCF+FIDE results, kind of like how the DBF format became the other standard. I could probably produce an unofficial specification for the format that would make it easier for you or others to implement it.
I truly believe we’re all in this together. This feels like a perfect example of how we could / should collaborate for everyone’s benefit.